Try to optimise char patterns

This commit is contained in:
Simonas Kazlauskas 2017-05-15 12:17:17 +03:00
parent 386b0b9d39
commit 41debc365e
3 changed files with 58 additions and 7 deletions
src
libcollections
libcore/str

View File

@ -195,30 +195,34 @@ fn bench_contains_equal(b: &mut Bencher) {
})
}
macro_rules! make_test_inner {
($s:ident, $code:expr, $name:ident, $str:expr) => {
($s:ident, $code:expr, $name:ident, $str:expr, $iters:expr) => {
#[bench]
fn $name(bencher: &mut Bencher) {
let mut $s = $str;
black_box(&mut $s);
bencher.iter(|| $code);
bencher.iter(|| for _ in 0..$iters { black_box($code); });
}
}
}
macro_rules! make_test {
($name:ident, $s:ident, $code:expr) => {
make_test!($name, $s, $code, 1);
};
($name:ident, $s:ident, $code:expr, $iters:expr) => {
mod $name {
use test::Bencher;
use test::black_box;
// Short strings: 65 bytes each
make_test_inner!($s, $code, short_ascii,
"Mary had a little lamb, Little lamb Mary had a littl lamb, lamb!");
"Mary had a little lamb, Little lamb Mary had a littl lamb, lamb!", $iters);
make_test_inner!($s, $code, short_mixed,
"ศไทย中华Việt Nam; Mary had a little lamb, Little lam!");
"ศไทย中华Việt Nam; Mary had a little lamb, Little lam!", $iters);
make_test_inner!($s, $code, short_pile_of_poo,
"💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩!");
"💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩!", $iters);
make_test_inner!($s, $code, long_lorem_ipsum,"\
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse quis lorem sit amet dolor \
ultricies condimentum. Praesent iaculis purus elit, ac malesuada quam malesuada in. Duis sed orci \
@ -253,7 +257,7 @@ Nam lectus enim, dapibus non nisi tempor, consectetur convallis massa. Maecenas
feugiat. Etiam quis mauris vel risus luctus mattis a a nunc. Nullam orci quam, imperdiet id \
vehicula in, porttitor ut nibh. Duis sagittis adipiscing nisl vitae congue. Donec mollis risus eu \
leo suscipit, varius porttitor nulla porta. Pellentesque ut sem nec nisi euismod vehicula. Nulla \
malesuada sollicitudin quam eu fermentum!");
malesuada sollicitudin quam eu fermentum!", $iters);
}
}
}
@ -288,6 +292,13 @@ make_test!(find_zzz_char, s, s.find('\u{1F4A4}'));
make_test!(rfind_zzz_char, s, s.rfind('\u{1F4A4}'));
make_test!(find_zzz_str, s, s.find("\u{1F4A4}"));
make_test!(starts_with_ascii_char, s, s.starts_with('/'), 1024);
make_test!(ends_with_ascii_char, s, s.ends_with('/'), 1024);
make_test!(starts_with_unichar, s, s.starts_with('\u{1F4A4}'), 1024);
make_test!(ends_with_unichar, s, s.ends_with('\u{1F4A4}'), 1024);
make_test!(starts_with_str, s, s.starts_with("💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩"), 1024);
make_test!(ends_with_str, s, s.ends_with("💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩"), 1024);
make_test!(split_space_char, s, s.split(' ').count());
make_test!(split_terminator_space_char, s, s.split_terminator(' ').count());

View File

@ -813,6 +813,7 @@ impl str {
/// assert!(!bananas.contains("apples"));
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn contains<'a, P: Pattern<'a>>(&'a self, pat: P) -> bool {
core_str::StrExt::contains(self, pat)
}
@ -900,6 +901,7 @@ impl str {
/// assert_eq!(s.find(x), None);
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn find<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize> {
core_str::StrExt::find(self, pat)
}
@ -944,6 +946,7 @@ impl str {
/// assert_eq!(s.rfind(x), None);
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn rfind<'a, P: Pattern<'a>>(&'a self, pat: P) -> Option<usize>
where P::Searcher: ReverseSearcher<'a>
{
@ -1057,6 +1060,7 @@ impl str {
///
/// [`split_whitespace`]: #method.split_whitespace
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn split<'a, P: Pattern<'a>>(&'a self, pat: P) -> Split<'a, P> {
core_str::StrExt::split(self, pat)
}
@ -1106,6 +1110,7 @@ impl str {
/// assert_eq!(v, ["ghi", "def", "abc"]);
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn rsplit<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplit<'a, P>
where P::Searcher: ReverseSearcher<'a>
{
@ -1152,6 +1157,7 @@ impl str {
/// assert_eq!(v, ["A", "", "B", ""]);
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn split_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> SplitTerminator<'a, P> {
core_str::StrExt::split_terminator(self, pat)
}
@ -1195,6 +1201,7 @@ impl str {
/// assert_eq!(v, ["", "B", "", "A"]);
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn rsplit_terminator<'a, P: Pattern<'a>>(&'a self, pat: P) -> RSplitTerminator<'a, P>
where P::Searcher: ReverseSearcher<'a>
{
@ -1247,6 +1254,7 @@ impl str {
/// assert_eq!(v, ["abc", "defXghi"]);
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn splitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> SplitN<'a, P> {
core_str::StrExt::splitn(self, n, pat)
}
@ -1294,6 +1302,7 @@ impl str {
/// assert_eq!(v, ["ghi", "abc1def"]);
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn rsplitn<'a, P: Pattern<'a>>(&'a self, n: usize, pat: P) -> RSplitN<'a, P>
where P::Searcher: ReverseSearcher<'a>
{
@ -1334,6 +1343,7 @@ impl str {
/// assert_eq!(v, ["1", "2", "3"]);
/// ```
#[stable(feature = "str_matches", since = "1.2.0")]
#[inline]
pub fn matches<'a, P: Pattern<'a>>(&'a self, pat: P) -> Matches<'a, P> {
core_str::StrExt::matches(self, pat)
}
@ -1370,6 +1380,7 @@ impl str {
/// assert_eq!(v, ["3", "2", "1"]);
/// ```
#[stable(feature = "str_matches", since = "1.2.0")]
#[inline]
pub fn rmatches<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatches<'a, P>
where P::Searcher: ReverseSearcher<'a>
{
@ -1415,6 +1426,7 @@ impl str {
/// assert_eq!(v, [(0, "aba")]); // only the first `aba`
/// ```
#[stable(feature = "str_match_indices", since = "1.5.0")]
#[inline]
pub fn match_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> MatchIndices<'a, P> {
core_str::StrExt::match_indices(self, pat)
}
@ -1457,6 +1469,7 @@ impl str {
/// assert_eq!(v, [(2, "aba")]); // only the last `aba`
/// ```
#[stable(feature = "str_match_indices", since = "1.5.0")]
#[inline]
pub fn rmatch_indices<'a, P: Pattern<'a>>(&'a self, pat: P) -> RMatchIndices<'a, P>
where P::Searcher: ReverseSearcher<'a>
{
@ -1737,6 +1750,7 @@ impl str {
/// assert_eq!(s, s.replace("cookie monster", "little lamb"));
/// ```
#[stable(feature = "rust1", since = "1.0.0")]
#[inline]
pub fn replace<'a, P: Pattern<'a>>(&'a self, from: P, to: &str) -> String {
let mut result = String::new();
let mut last_end = 0;

View File

@ -429,7 +429,33 @@ impl<'a> DoubleEndedSearcher<'a> for CharSearcher<'a> {}
/// Searches for chars that are equal to a given char
impl<'a> Pattern<'a> for char {
pattern_methods!(CharSearcher<'a>, CharEqPattern, CharSearcher);
type Searcher = CharSearcher<'a>;
#[inline]
fn into_searcher(self, haystack: &'a str) -> Self::Searcher {
CharSearcher(CharEqPattern(self).into_searcher(haystack))
}
#[inline]
fn is_contained_in(self, haystack: &'a str) -> bool {
if (self as u32) < 128 {
haystack.as_bytes().contains(&(self as u8))
} else {
let mut buffer = [0u8; 4];
self.encode_utf8(&mut buffer).is_contained_in(haystack)
}
}
#[inline]
fn is_prefix_of(self, haystack: &'a str) -> bool {
CharEqPattern(self).is_prefix_of(haystack)
}
#[inline]
fn is_suffix_of(self, haystack: &'a str) -> bool where Self::Searcher: ReverseSearcher<'a>
{
CharEqPattern(self).is_suffix_of(haystack)
}
}
/////////////////////////////////////////////////////////////////////////////